As is known in the art, because of increased mobility of people throughout the world increasing numbers of people are now residing in countries where their native language is not widely spoken. For example, immigrants who are non-native speakers of the English language are entering the United States. Many of these immigrants are proficient in written English however, because of poor pronunciation due to accents they are unintelligible when they speak. The problem of lack of intelligibility in non-native speakers of English may be particularly recognized at universities which have immigrants working as teaching assistants (TAs). This is particularly true in the science and engineering fields. The need has thus arisen for improved accent reduction services.
Courses which teach English as a second language, for example, have been used for this purpose. A widely accepted aspect of current theory in such courses includes training in self-monitoring. Self-monitoring in this context includes graded exercises in listening to oneself speak, while focusing on a particular production feature. Self-monitoring in the sense of listening to oneself during the act of speaking, while useful, is not particularly easy.
In traditional audiotape lessons the act of speaking may be separated from the act of listening to one's own speech. While self-recording is possible with most instructional audiotapes, the learners very seldom actually listen to their recordings. Rewinding the tapes is awkward and results in an unacceptable delay. Listening to the self-recordings is not an integral part of using the tapes, but rather an additional feature which has the result of slowing down the time to complete the lesson and thus to users this feature appears to be "extra work." Furthermore, in such systems user control of the timing and sequencing of exercises and the number of repetitions is often minimal. This results in a passive learning experience. Lastly, the tasks to be performed in the audiotape lessons generally focus on the learning of syntactic structures through exercises such as substitution drills, with improvement of pronunciation being a subsidiary goal.
While courses which teach immigrants a second language may be the natural context in which immigrants may improve their pronunciation, such courses are usually oversubscribed, the language is taught at too basic a level, and it is difficult to address the pronunciation problems of specific individuals. Furthermore, instructors in such courses often have little training in language pathology or articulation therapy.
Speech-language pathologists, having an educational background in phonetics and articulation therapy, are well-equipped to provide this service. Speech-language pathologists are appropriate instructors for immigrants and some speech-language pathologists have developed specializations in this area. Nevertheless, the need for such services seems to be growing more rapidly than the services available.
One solution to this problem has been to provide language learning systems which use visual displays to aid speakers in identifying problems in their speech. One such system employs a spectrographic display of speech signals which may be used to train speakers to correct articulation errors. However, several problems exist with the spectrogram technique such as the fact that not all articulatory features may be visible in the spectrograph, the use of spectrograms requires both considerable knowledge on the part of the clinician and extensive guidance to the client. Furthermore it may be difficult to generalize the acquired articulations to spontaneous speech. Moreover, while visual feedback is useful in identifying problems, the user eventually relies on the auditory system to improve pronunciation.